Sequence of Returns Risk: Why the Order of Gains and Losses Matters

Why order does not matter until money moves

If you invest a lump sum and never add or take anything out, the order of returns makes no difference. Multiply the yearly growth factors together and you get the same final number whatever the sequence.

Withdrawals change that. When you take money out in a bad year you sell assets at low prices, and those shares are no longer there when the market recovers. Early losses therefore do more damage than late ones, because the withdrawals taken during the slump shrink the base that has to recover.

Contributions do the opposite. A saver adding money buys more shares after a fall, so for them early losses can help. We explore that in our guide to dollar-cost averaging versus lump-sum investing.

A worked example

The figures are hypothetical. Take ten annual returns: +20%, +16%, +14%, +12%, +10%, +8%, +4%, −2%, −10% and −12%. Their arithmetic average is 6% a year, and the compound average is about 5.5%.

Both start with $1,000,000 and withdraw $40,000 at the start of each year, a 4% initial withdrawal. To keep the arithmetic clear we do not adjust for inflation.

Look at Retiree B's portfolio along the way. After the first year it is $844,800. After the third it is $670,634, about a third below the starting value, while withdrawals are still continuing. It then grows again, but from a smaller base. Retiree A's balance peaks at about $1.78 million in year seven.

The bottom row shows the mirror image. An investor adding $40,000 a year does better when the losses come first, by the same $300,098.

• Retiree A gets them in that order, strongest years first.
• Retiree B gets them in reverse, with the losses first.

What the research shows

Bengen's 4% study. William Bengen's 1994 article in the Journal of Financial Planning tested withdrawals against historical US stock and bond returns. He found that a first-year withdrawal of 4%, followed by inflation-adjusted withdrawals, should be safe, and it survived no less than about 33 years in all the retirement periods he examined. A later review in the same journal (November 2023) describes the result as a maximum initial withdrawal rate of 4.0% for a portfolio of US large-cap stocks and intermediate-term Treasuries over at least 30 years, and notes that the worst-case cohort in its analysis was the 1973 retirement year. In other words, the retirees who started in 1973 faced the hardest sequence in that analysis.

Morningstar's analysis. Morningstar's State of Retirement Income research defines the risk as the chance that losses early in retirement jeopardise a retiree's ability to sustain spending to the end of life. Its authors, Amy Arnott, Christine Benz, Tao Guo and Jason Kephart, found that nearly 70% of the "failures" in their simulations involved retirees whose investments had lost value by the end of year five of retirement. They also report that an all-equity portfolio supported a starting withdrawal rate of 3.1% at a 90% success probability over 30 years, and that diversifying toward bonds raised the safe starting rate, to 3.7% for balanced allocations.

In Morningstar's 2022 edition of the report, the authors put the safe starting withdrawal rate at 3.8% for a 50/50 stock and bond portfolio, a 30-year horizon, a 90% probability of success and fixed inflation-adjusted withdrawals. The figures above come from different editions and assumptions, so they are not directly comparable. The point is that the number moves with market conditions and assumptions, which is a reason to treat any single "safe rate" as a starting estimate.

What you can do about it

None of these removes the risk. Each trades one cost for another.

This article is general information, not personal advice. The examples use hypothetical returns and exclude taxes and inflation.

1. Hold a mix that cushions early losses. Morningstar found safe starting withdrawal rates rose as portfolios moved away from all-stock toward bonds. The cost is lower expected growth over time.
2. Keep spending flexible. Morningstar's authors recommend flexible withdrawal approaches and note that retirees facing a downturn can dial back spending temporarily. Even a modest cut in the first bad years reduces the number of shares sold at low prices.
3. Start with a lower initial rate if your spending is rigid. Bengen's 4% and Morningstar's 3.1% to 3.8% show how much the answer depends on mix and assumptions.
4. Know the danger window. In Morningstar's simulations the risk is concentrated in the first years. Check the portfolio, and your spending, more carefully then.
5. Test your own numbers. Run your plan with a bad first five years and see what it takes to survive it. A plan that only works on average has not been tested.